June 6, 2013:
Domino Search was posted several years ago to help solve a
puzzle type I found in the book
Logical Puzzles, Chartwell Books. (BTW, Amazon currently has used
copies of the book starting at $4.00 shipped at this link.) In this
program, the viewer is given an array
of numbers representing dots on the 28 standard dominos, but without the domino
outlines. The player's job is to replace them. The program was only
partially successful because it could generate and solve random arrays it
generated, it also allowed users to enter data and supported user play, but it
didn't know how to solve those puzzles. The first step toward fixing this
is a separate program which accepts and solves user submitted arrays and
saves the array and solution in a file for input to Domino Search.
Domino Search Version 5, and Define Domino
Arrays Version 2 posted today are implementations of this strategy.
Perhaps one day I'll get around to rolling them back into one program, but not
May 25, 2013: It has been a busy month with 2 grandchildren graduating from college, spring clean up of extensive winter damage and the always stressful planning for a big family vacation in June.
I did find time to update the 15 Graphic Effects program to Version 2 in our Delphi Techniques section. The update improves the efficiency of the "Magnifier" effect and has better scaling for demo images loaded. The original program was written 10 years ago by bright 16 year old Ivan Sivak. I emailed him the other day asking for a status update but haven't heard back yet.
Table at left shows an original photo and 4 of
the 15 special effects: Brightness/Contrast, Gray Scale, Blend, and
May 11, 2013: The original version of our "Copy Folder" utility program performed operations like copying all files in folder "ProgramA" to a new folder "ProgrtamA_Test". A viewer pointed out it did not work well for his intended use: i.e. copy folders "ProgramA", then "ProgramB", and then "ProgramC" into a folder named "Backups". In other words, copy the selected folder name as well as the files in that folder. It seemed like a reasonable request to me, so CopyFolder Version 3.2 posted today adds a checkbox: "Include selected input folder record in output" to do just that.
May 7, 2013: Here is a program which solves a problem discussed by computer science pioneer Donald Knuth in 1977 in a magazine article "Are Toy Problems Useful?" in 1977. He was disputing the argument made by another mathematician that the answer was "No". I agree with Knuth, that, like story problems in math textbooks, any problem which improves problem solving skills is useful even if the answer will not directly advance one's professional career. This problem presented by Knuth met the "useful" criteria for me. "Given an exponent, N, find all numbers which are equal to the sum of the Nth powers of their digits." For example for N =4: 1634 = 14 + 64 + 34 + 44 (= 1 + 1296 + 81 + 256 = 1634). Knuth Toy Problem has the results of my investigation, three methods which, for N=10, find the unique solution in 30 minutes, or 2 hours, or 2 seconds. A rare personal Eureka moment! The explanation for the best result is behind a button in case any programmer wants to discover it on their own J.
April 23, 2013: An update to make Know, Don't Know Version 3.1 today added a second "Walk-through" page, this one interactive, taking any sum and product and stepping through the analysis from each professor's point of view. This was motivated by a viewer who doubted that the validity of the second solution for the 500 upper limit case. To eliminate the possibility that of a bug in the original solver code, I needed to step through as the Professors would have done. The size of the numbers makes this an order of magnitude harder than the first "Walk-through" and justified writing the additional code.
April 20, 2013: A viewer wrote last week regarding my "Know - Don't Know" program which analyzes a logic puzzle involving two people (usually professors), one given the sum of two numbers and the other given their product. By exchanging non-numeric messages about what they know or do not know, they both manage to find the numbers. It is often called the "Impossible Problem" because it seems that that should be the case. My original versions concentrated on finding the numbers without knowing either the sum or the product, but glossed over how the professors, particularly the one knowing only the sum, might have solved the problem. Know, Don't Know Version 3 posted today adds a "Walkthrough" page to the the program describing the thought processes of each professor at each exchange leading to them both finding the solution. It helped give me, and hopefully Charles and others, a deeper understanding of the problem and its solution.
April 14, 2013:
Brain Game calendar puzzle solver was posted today. If you are a puzzleist
and need help finding the solution to this or a similar puzzle, download the
executable version of Expressions
From Integers. If you are a programmer or interested in how a computer
program might solve this puzzle and others like it, you might enjoy browsing the
text on the web page and/or the downloaded source code.
April 6, 2013: One more fairly large conversion of Delphi to Lazarus posted today; the 6 individual word based programs (Crossword Helper, Decrypt, Scrambled Pie, Spellbound, Unscramble, and Word Ladder) plus the wrapper, Wordstuff 3, that links to any of them all wrapped up in a single zip file. I have added some notes to the Lazarus Revisited page. I've decided to post Lazarus notes there with the latest additions in red.
April 2, 2013: DFF Newsletter #67 was sent yesterday to subscribers. Newsletters are primarily a listing of the "What's New" items for the quarter. The motivation was to update those who find the site interesting but do not visit routinely.
The Index of Lazarus conversions is now available as a link from the Lazarus Revisited page. Also, I just added the file of adage candidates which was missing from the Adage Anagrams program download posted in January.
March 28, 2013: We're embarking on a grand experiment today - converting DFF Delphi programs to Lazarus/FPC. Lazarus is a frontend IDE (Integrated Development Environment) for FPC (Free Pascal Compiler). It knows about Delphi and, in it's current form, does a pretty fair job of converting Delphi source. For someone wanting to use DFF programs on a non-Windows platform or who wants to modify DFF code but does not have access to Delphi, it's a no-cost approach worth a shot. Lazarus Revisited introduces the process I used to convert most of the widely used DFF Library unit and a dozen or so programs using them.
March 14, 2013: A new Utility program, Bulk Find & Replace, was posted today.
Like most of the programs in the Utilities section of DFF, it was motivated by a specific problem that I had. While at it though, I made the program generalized with these options and features
Let me know if you find it useful (or especially if you find bugs!) .
March 3, 2013: It didn't take long to come up with Interesting 2013 Version 2.1 which takes advantage of the "prime numbers only" search to make searches 100 times faster than the previous version. Also my new candidate for the smallest number requiring six terms is 11. If we allow term sizes to be up to two times larger than the target number then the smallest number requiring six terms 2 (= 32 + 32 -22-22-22-22).
March 1, 2013: On January 4th I posted a program investigating two erroneous claims made about interesting features of the year 2013. A viewer recently pointed out a problem with my proof that 2013 was not, as claimed, the smallest number which required 6 terms to be expressed as the sum or difference of prime numbers squared. I forgot to select only prime numbers, but the claim is still untrue in the corrected version. 432 + 132 - 32 + 22 does the job with 4 terms. In fact, I believe that the 15 is the smallest number requiring 6 prime squared terms if individual term values are limited to the target number value. If you need help finding expression for 15, search button for Case 2B in Interesting 2013, Version 2 can help.
February 20, 2013: It took a while, but a user
finally found a bug with the 3-player option in our Four In a Row game.
Retracting (undoing) a move switches forward to the next rather than back to the
previous player. In the 2-player options it didn't matter of course but,
with 3-players, trying to undo a move advances to the next player and the "Reset"
button is the only way out. Four In A Row Version 2.3 posted today
fixes the problem.
February 17, 2013: If we break an integer, N, into smaller integer parts which add up to N, the sets of numbers are called Integer Partitions of N. They have been the subject of study for a few hundred years and are still studied today. A Google search will lead you down the path of discovery as far as you care to travel. My Integer Partition Test program, Version 2 was posted today. In addition counting and generating partitions for integers up to 375, it will generate partitions with a specific maximum value or a specific number of parts. Interestingly, for a given integer K, the number of partitions of each of the types is the same. Although it is generally not feasible to list all partitions of even a medium size integer (there are 190 million ways to partition the integer 100), the program can now calculate the partition for any position (rank) for any input number up to 375. So if you want to know the millionth, or billionth partition of 200, have at it!
February 7, 2013: PosExTest is a program in our Delphi techniques section that defines and tests a substitute for the Delphi substring search function, PosEx. PosEx was not available in versions before Delphi 7.
February 4, 2013: I received a note today from Dieter Stein, creator of the Paletto game I cloned last month. He liked the program but pointed out that the initial game board setup should not allow adjacent balls or tokens of the same color in the vertical or horizontal direction. I corrected my program and reposted Paletto Puzzle Version 2.1 today.
February 1, 2013: Another puzzle from our 2013
"Brain Game" page-a-day calendar. I worked on this one for about 10
minutes before deciding that a 10 minute program would be more fun and
more likely to guarantee a solution. It took more like 30 minutes
to write No_3_In_A_Row, but it does
find the two solutions in less than a second.
The What's New Archives
(Click to expand)
Copyright © 2000-2013, Gary Darby
All rights reserved.